Table 24 indicates that some has been removed from Tor tugas sea-water, as compared with other sea-water, and to a greater extent from Key West sea-water. In other words, the precipitation observed by Vaughan is not due to a greater amount of calcium in Tortugas or Key West sea-water, but to local conditions which cause the precipitate to form.
According to the law of mass-action, in a saturated solution of in sea-water at constant temperature, salinity, etc., Not all of the calcium is, however, in the form of and Ca • ', for some is undissociated and The chlorides and sulphates are constant, but and change with the total content of the sea-water. But I have shown (McClendon, 1917b) that if the alkaline reserve remains constant the total of the sea-water (within limits found in nature) varies inversely with the pH ( = —log. H' concentration). Hence the de termination of the pH may be substituted for that of the total The determinations I have made of the water of the Pacific and North Atlantic showed the pH to vary from about 8.1 to 8.25 and those of Dr. A. G. Mayer in the Pacific showed only a little wider range (table 11). Earlier observations at Tortugas gave the same range, but my more extended observations in this summer of 1917 show the inadequacy of a few determinations. The pH is influenced by plant and animal life and rises at Tortugas to 8.35 during the day over well lighted bottoms rich in vegetation, and falls to 8.18 during the night. It may be said, therefore, that conditions in shallow water over eel grass or other seaweed or corals (with symbiotic alga') favor the pre cipitation of The question arises whether the occasional high pH of Tortugas sea water is sufficient to explain the precipitation of or whether nuclei for the separation of the solid phase are necessary. A large amount of may be added to sea-water without causing a pre cipitation. If the pH is increased by the addition of NaOH, the result depends on the speed at which the alkali is added. If the NaOH is added suddenly in the form of a strong solution, colloidal precipitation membranes form about the drops and if the membranes are broken by shaking or stirring, a great mass of is included in the pre cipitate. If a very dilute solution of NaOH is added very slowly, possibly contaminated with is precipitated. The
exact pH at which precipitation first occurs can not be determined by this method, as the first precipitation occurs before the solutions are mixed and the crystals thus formed serve as nuclei for further precipita tion. If Tortugas sea-water is kept in glass bottles, precipitation occurs on the glass while the pH of the water is within the natural limits, but the pH at the glass surface is higher, due to solution of glass.
Although the pH at which precipitation would occur without nuclei for the separation of the solid phase may be practically impossible to determine, the final equilibrium with an abundance of nuclei is not a difficult problem. Calcite and aragonite crystals to serve as nuclei were produced by the methods of Johnston, Merwin, and Williamson. The crystals were examined under the microscope and tested with cobalt-nitrate solution. These observations, together with the mode of preparation, leave little doubt that the crystals actually were calcite and aragonite. Under the microscope an occasional calcite crystal could be found among the aragonite crystals, but the number was not sufficient to affect the cobalt-nitrate test. These calcite crystals seemingly grew slightly during the experiments, but apparently no new ones were formed. To determine the equilibria, crystals were mechanically stirred or shaken with sea-water in "nonsol" flasks, 6 to 14 hours at 30'; then the pH and alkaline reserve were determined.
The results of agitating 100 grams of calcite crystals with a liter of sea-water until equilibrium was approximately reached also appear in figure 7. The alkaline reserve is measured on the ordinate and the pH on the abscissa. This shows that sea-water of the surface of the oceans of the whole world is supersaturated in respect to calcite. We may therefore conclude that suitable nuclei for the precipitation of calcite are absent or deficient in number. The solubility of crystals varies inversely with their size, but after they have attained sufficient size to be readily examined with low powers of the microscope further increase has an unappreciable effect on solubility. But such crystals, if present, would rapidly gravitate to the bottom of the sea; hence the absence of nuclei for precipitation of calcite is what one might expect.